neurovirulence test of types 1, 2 or 3 live … · who sop for neurovirulence testing of oral polio...

STANDARD OPERATING PROCEDURE

NEUROVIRULENCE TEST

OF TYPES 1, 2 OR 3 LIVE ATTENUATED POLIOMYELITIS

VACCINES (ORAL) IN TRANSGENIC MICE SUSCEPTIBLE

TO POLIOVIRUS

Version 6

(2012)

WHO SOP for neurovirulence testing of Oral Polio Vaccine using TgPVR21 transgenic mice Version 6 (2012)

of 39

Table of contents

Page

1. Introduction 3

2. Transgenic mice 4

3. Titrations of vaccines in cell culture and preparation of inocula 4

3.1 Titrations of vaccines in cell culture

3.2 Preparation of inocula

4. Animal procurement and randomization 5

5. Intraspinal inoculation and clinical observations of mice 5

5.1. Intraspinal inoculation

5.2. Clinical Observations

6. Statistical analysis and decision model 8

Appendices

Appendix 1. Randomization procedures for OPV neurovirulence tests

in transgenic mice (TgmNVT)

10

Appendix 2. Recommended training program for the transgenic mouse

neurovirulence test (TgmNVT) and for the training of new inoculators and

maintenance of competence of qualified inoculators

21

Appendix 3. Recommended implementation process for the transgenic mouse

neurovirulence Test (TgmNVT)

26

Appendix 4. Statistical analysis and decision model for testing Serotype 1, 2 or

3 OPV vaccines in transgenic mice.

29

Appendix 5. Monitoring of data from the transgenic mouse neurovirulence test

(TgmNVT).

36

Authors

37

References

37


of 39

1. Introduction

The monkey neurovirulence test for monovalent oral polio vaccine (OPV) bulks has been used, in

various formats, for over 50 years. While this in vivo test has remained an important component

of OPV quality control, the use of non-human primates has become increasingly difficult in many

parts of the world. In the 1980s transgenic mice carrying the human poliovirus receptor were

developed, and in 1992 WHO initiated a project to evaluate the suitability of transgenic mice for

OPV neurovirulence testing. Initial studies with Serotype 3 monovalent bulks were encouraging

and a series of international collaborative studies were carried out in the late 1990’s to validate

the design and operation of a transgenic mouse neurovirulence test (TgmNVT), first for Serotype

3, and subsequently for Serotypes 1 and 2 vaccines. The transgenic mouse test for all three

poliovirus types is now part of the WHO recommendations for the production and control of

OPV(1).

The test is based on behavioral ("clinical") observations of groups of TgPVR21 transgenic mice

inoculated intraspinally with either a test vaccine or a standardized WHO reference material (2).

Using the standardized procedures described in this SOP mice are scored as "normal" or

"paralysed," the differences between groups analysed statistically, and test vaccines that are

significantly more neurovirulent than the WHO reference vaccine are rejected. The experimental

design requires 32 mice (16 male and 16 female) at each dose level for each vaccine. Up to two

test vaccines may be included in a single test, which may be split over two days. A total of 128

mice (one test vaccine) or 192 mice (two test vaccines) are required. The design is balanced over

vaccines, doses, gender, and where appropriate, days. Randomization of the allocation of mice to

vaccines and doses, the cage locations, and the order of inoculations is essential as is the blinding

of these allocations to both inoculators and scorers until the completion of the test. A detailed

method of statistical analysis is defined, along with associated validity tests. In an analogous

manner to the monkey NVT, there is the possibility of retesting built in to decision procedure.

The Standard Operating Procedure presented here describes the currently recommended

procedures for the conduct of this test for all three oral polio vaccine serotypes. It also confirms

the recommendation from 1999 (1) that a standardized approach be taken for the qualification of

institutions that wish to adopt the test for use in quality control. The recommended training

program for the intraspinal inoculation technique is described in Appendix 2. After completion of

the training program laboratories are then required to complete a standard implementation

procedure to demonstrate their overall competence to perform the TgmNVT. Only after

completion of the training and implementation process should a laboratory consider the use of

batch release testing using the TgmNVT. The final decision relating to whether a laboratory can

or cannot use the TgmNVT for the control of poliovirus monovalent bulks used in OPV

production resides with the National Control Authority (1).

Additional information on reagents, training, and contacts relating to the TgmNVT are available

on the WHO web site www.who.int/biologicals or by contacting WHO at:

Coordinator, Quality, Safety and Standards

Essential Medicines and Health Products

World Health Organization

20, Avenue Appia

CH-1211 Geneva 27, Switzerland


of 39

2. Transgenic mice

Only an approved strain of mice obtained from a WHO approved source should be used. At

present the only approved strain is the TgPVR21 mouse line developed in Japan (1). Procedures

and standards for maintenance and distribution of transgenic mice susceptible to human viruses

described in national or international guidelines should be observed (3), and particular notice

should be taken of the containment requirements that will be necessary once endemic

transmission of wild poliovirus strains has been successfully ended (4). The animal facility

should be approved for this purpose by relevant national authorities before work with TgPVR21

commences, with copies of the documentation relating to the approval of the facility made

available to the supplier and/or shipping agent if required.

3. Titration of vaccines in cell culture and preparation of inocula

3.1 Titrations of vaccines in cell culture

The recommended method for the titration of viruses is outlined in the WHO Recommendations

to Assure the Quality, Safety and Efficacy of Live Attenuated Poliomyelitis Vaccine (oral) (1).

Briefly, this method is based upon a determination of the cell culture infectious dose (CCID50) in

Hep2C cell cultures. For estimation of the titre of any sample, replicates should be performed to

obtain a precision of ±0.3 log10 CCID50/ml or better for the 95% confidence limits of the mean.

The dilutions of vaccine may be made in advance and aliquoted in multiple containers which

should be stored at ≤ –70° C.

Titres of the diluted samples should preferably be determined by titration before the test is started

to ensure that the doses to be given are correct. The required doses for each virus serotype tested

against the WHO(SO+2) reference virus are:

Upper Dose

(log10 CCID50/5µµµµl)

Lower Dose

(log10 CCID50/5µµµµl)

Type 1 2.75 1.75

Type 2 6.0 5.0

Type 3 4.5 3.5

One container per dose should be thawed on the day of the inoculations and used on that day

only. The diluted vaccine containers should be stored on ice during the inoculation procedure.

Residual inocula may be aliquoted and stored frozen until the back-titrations are carried out.

Three replicate back-titrations of the residual viral inocula should be performed after the

injections of mice are completed to confirm that the required doses were tested. An aliquot of

diluted vaccine stored continuously at ≤ –70°C may also be titrated in parallel, particularly if the

diluted vaccines were not tested prior to inoculation of the samples.

As clinical manifestations of neurovirulence in TgPVR21 mice are strongly dose-dependent,

extreme care should be taken to ensure that the dilution of vaccines is performed as accurately as

possible, and for the test to be valid, the mean titer of the residual inocula should be within ± 0.3

log10 CCID50/ml of the target dose.

3.2 Preparation of inocula


of 39

The vaccine and reference preparation should be diluted with a suitable diluent, such as those

used in monkey neurovirulence testing. Eagle's minimum essential medium (EMEM) containing

0.14% bovine albumin and 0.22% sodium bicarbonate, and Earle's balanced salt solution

containing 0.5% lactalbumin hydrolysate have been successfully used by the National Institute

for Biological Standards and Control (NIBSC), UK, and the U.S. Food and Drug Administration

(FDA), respectively. Diluents have been shown to be important for obtaining consistent results,

and if other diluents are chosen, they should be validated for equivalence to these reference

diluents prior to use.

4. Animal procurement and randomization

TgPVR21 transgenic mice should be ordered to arrive at the testing institution at the age of five

to six weeks and the randomization completed as soon as possible after the arrival of the mice.

The mice should be allowed to recover from the shipping and randomization procedure for at

least seven days before inoculation. To minimize the risk of cross-contamination it is preferable

that animals should be inoculated within fourteen days of arrival at the facility. Aggressive mice

should be housed in individual cages, kept separated throughout the duration of the test, and this

exception recorded in the test record. Randomization is an important element of the TgmNVT

and it is essential that the correct randomization procedure is followed. The procedures in

Appendix 1 describe the method for randomizing mice by dose, cage location, and order of

inoculation.

5. Intraspinal inoculation and clinical observations of mice

The local operating procedure should ensure that a clear record with traceability of the individual

mice, cages, vaccine/reference and dose allocation is kept. However, the procedure should

ensure that this information is unknown (blinded) to the inoculator and clinical scorer until after

the completion of the full test.

5.1 Intraspinal inoculation

Prior to the inoculation mice should be marked in such a way that each individual in a cage can

be identified. The use of differently colored indelible marker pens to streak the hair or tail has

been found to be a successful and non-injurious method. The backs of the mice should be shaved

or depilated to facilitate inoculation. Mice should then be anesthetized with an anesthetic mixture

prepared on the day of inoculation; an example of an anesthetic is a mixture of ketamine (10

mg/ml final concentration) and xylazine (0.4 mg/ml final concentration) in physiologic saline

which is injected intraperitoneally in volumes of 0.3 to 0.4 ml depending on the weight of the

animal. Mice should be observed under anesthesia until they are immobile and lose pedal

reflexes. The skin is then disinfected with alcohol or an equivalent disinfectant, gently lifted to

form a “tent,” and a 1 to 2 cm longitudinal incision is made over the thoracic-lumbar regions of

the spine. For inoculation, the mouse is placed prone with its abdomen draped over a test tube or

other suitable object approximately 2 cm in diameter. Inoculations are performed using a

microsyringe with a 25-µl volume and a custom-made 33-gauge needle, 1.25 cm long with a

bevel angle of 15° and an external supporting 25-gauge cannula. Syringes and needles are

sterilized by autoclaving for 20 minutes at 121°C. Before autoclaving the syringes must be

treated with silicone solution (1 part of 360 medical grade silicone in 9 parts of 1,1,1,

trichloroethene) to ensure smooth movement of the plunger.

The needle is inserted between the spinous processes of the last thoracic and the first lumbar

vertebrae, directed cephalad at an angle of approximately 70° to the long axis of the spine, and


of 39

advanced into the substance of the spinal cord. The needle is then pulled back for half of the

length inserted into the spinal cord, turned 45° from the midline, and advanced again into the

anterior gray matter or anterior horn. If the needle is inserted into the correct position, one or

both hind legs should jerk. If there is no jerk, the needle should again be partially withdrawn, the

angle of insertion slightly adjusted, and the needle again advanced until a leg jerk is observed. If

the second insertion produces no jerk, the needle should be withdrawn completely and a new

point of insertion selected. The number of attempted needle insertions should not exceed three.

Once a jerk reaction is observed, 5 µl of the sample should be injected slowly over a period of 2

to 5 seconds to minimize pressure caused by the entry of the injection volume while avoiding

mechanical trauma caused by needle movement. The injection may be conducted by either one

or two persons depending on whether one person can simultaneously position the needle and

push the syringe plunger with ease. Where this proves difficult, or where the number of animals

with mechanical inoculation trauma is high, a second person should be employed to push the

plunger while the needle is held firmly in place by the first. During injection of the test material

mice should exhibit jerks of the hind legs, and in some cases, twisting of the rear half of the body

may be observed. Following inoculation the skin incision should be closed with a suitable tissue

adhesive or by an alternative means, and ophthalmic ointment applied to each eye to prevent

desiccation of the cornea. It is recommended that the same inoculator performs the injections for

a whole test. If injections are performed over multiple days the same inoculator should perform

the injections over the whole day.

5.2 Clinical Observations

Animals should be observed for neurological signs of disease at least once per day for 14

consecutive days after inoculation and the clinical stage recorded for each observation. An

additional observation 8 hours after inoculation may be conducted optionally, particularly for

tests of Serotype 1 vaccines (see below). Persons inoculating mice or assigning clinical stage

scores are to be blinded (unaware) of which test vaccine or dose has been administered to an

animal to prevent observational bias. Four stages of clinical symptoms are defined based on

characteristic motor signs:

Stages Physical signs

Normal Grips the edge of the

cage

Walks normally on

the grid and on a flat

surface

Full ability to move

limbs forward

Weak Unable to grip the

edge of the cage Walks normally on a

grid or flat surface Full ability to move

limbs forward

Paresis/Partial

paralysis Unable to grip the

edge of the cage

Limb falls through

the grid more than

once while walking

steadily forward

and/or toes curl

repeatedly while

walking on a flat

surface

At least a partial

ability to move limb

forward

Paralysis Unable to grip the

edge of the cage No use of limb on

grid or flat surface Inability to move the

limb forward

The above 4 stages may be noted using a code devised by the observer but the code should be

consistent and clearly identified on the score sheet.


of 39

For clinical scoring the examination of each mouse shall follow a step by step process:

1. The general state of the mice in the cage is observed to note possible morbidity, sickness,

or injury due to aggressive behavior.

2. Weakness in the hind legs (where it often appears first) is tested by holding the mouse by

the tail head down with its feet against the vertical wall of the cage. When a normal

mouse is then raised, both hind limbs will instinctively grip the upper edge of the cage

wall firmly. Weakness is defined by a failure of one or both hind feet to grip the cage

edge. Weakness of the front limbs can be assessed by placing the front paws of the mouse

on a cage lid or other similar surface and checking for resistance (firm gripping) when the

animal is gently pulled or lifted by the tail. If all limbs are able to grip, the mouse can be

scored as normal. If any limb is unable to grip, or if the score is questionable, the

observations described in step 3 should be carried out.

3. The mouse should be observed walking. During the observation the mouse may be held

by the tail but should be able to move in an unrestricted manner. A deficit in moving one

or more limbs forward, a failure to raise the heel while walking, or repeated curling-under

of the toes as the limb moves forward should be scored as paresis. Paralysis is scored

when one or more limbs are dragged as the animal moves forward. If necessary the

mouse may be made to walk on a grid (cage lid or similar surface) to aid in distinguishing

the difference between weakness and paresis. If no limb passes through the grid as the

animal is moving forward it can be scored as weak.

In borderline cases where it is difficult to decide between clinical stages, the observer should

assign the lower score. Scores should be recorded for each observation. The score sheet should

also note in which limb the symptom is observed to insure consistency in reporting.

In tests on serotypes 1 and 3 a steady progression is generally observed through weakness to

paresis and paralysis. For serotype 2 however mice may show partial recovery where paralysis

reverts to paresis or paresis to weakness. It is therefore essential that daily observations be

conducted.

Only the advanced motor disorders of paresis and paralysis can be considered as reliable stages

for determining the acceptability of a test vaccine. Therefore at the end of the experiment,

animals should be scored as either Normal or Paralyzed (0 or 1). Paralyzed (1) animals are

defined as those that:

1) Showed a paralysis clinical score in any limb, or

2) Showed a paresis clinical score on 2 consecutive days. Under normal practices, scoring is

conducted every 24 hours, and thus two consecutive days will be equivalent to two

consecutive scores.

For ethical reasons, animals which are scored as paresis on 2 consecutive days or paralysis for

one day may be euthanized. Death by euthanasia in this case should be noted on the data sheet in

addition to the clinical stage and the mouse will be scored as paralysed (1) in the final score. All

other surviving mice are to be euthanized on day 14.

Animals exhibiting any of the following criteria should be recorded but should be excluded from

the final analysis:


of 39

1) Death due to anesthesia, normally resulting in the animal not recovering from inoculation;

2) Death due to inoculation trauma, including euthanasia of severely traumatized animals;

3) Animals that show traumatic paresis/paralysis that appears within 24 hours after the

inoculation where the paralysis does not progress. This is not applicable in cases where

the animal recovers after the initial trauma. In this case the animal is scored according to

the normal procedure.

4) Death as a direct result of fighting or injury;

5) Other non-identified causes of death where a relationship with poliovirus clinical

progression is not evident;

If more than 5% of animals are excluded for the reasons noted above the test should be repeated.

It should be noted that with Serotype 1 a more rapid onset and progression of paralysis compared

to the other two serotypes has been observed infrequently, making a differentiation between

traumatic injury and poliovirus-related clinical signs more difficult. To monitor this, at the

discretion of the testing institution mice may be additionally scored at 8 hours and 24 hours after

inoculation so that this rapid progression of symptoms can be identified. This can also be

performed with tests with Serotype 2 and 3 vaccines. Under blinded conditions this would require

scoring all animals in the test. When animals die without any apparent cause, depending on the

physical condition of the mouse body and the presence of the observer at animal death or

immediately after, a post-mortem examination should be conducted by a qualified person to rule

out poliovirus-related causality. Additional controls such as histological examination of the

spinal cord or brain stem, or PCR testing of heparinized blood for the presence of poliovirus may

be performed at the discretion of the testing institution. If frequencies of >5% inoculation trauma

occur, a thorough investigation into the possible causes should be conducted and corrective

actions taken.

6. Statistical analysis and decision model

The analysis of the TgmNVT is based on a comparison of the proportion of mice paralysed at

each dose of the reference and test vaccines. The current WHO reference vaccines are

representative of vaccine lots studied in clinical trials and found acceptable for the purpose of

immunization. If a test vaccine is equally or less paralytic (neurovirulent) than the reference

vaccine it is considered acceptable. If it is more paralytic it is unacceptable. In the event where a

clear statistical decision is not possible, the possibility of a retest is included in the procedure.

The TgmNVT is analysed by a logistic model relating the observed paralysis proportion to the

vaccine, the dose, and the gender. The relative neurovirulence of the test vaccine compared to the

WHO reference vaccine is given by the “odds ratio,” a number that represents how much more

likely it is for paralysis to occur with the test vaccine than with the reference vaccine. An odds

ratio of one indicates that the test vaccine and reference vaccine have equal neurovirulence. High

values of the odds ratio indicate that a test vaccine is more neurovirulent than the reference

vaccine. The log of the odds ratio (LOR) is calculated, and compared to a pre-determined L-value

(analogous to the C-values in the monkey NVT). L1 and L2 values are the acceptance and

rejection limits calculated from the probabilities that differences in scores are due to chance

alone. If the observed LOR is less than the L1 value, the vaccine passes. If the LOR is above the

L2 value the vaccine fails. If the LOR falls between the L1 and L2 values then a retest is required

to make the decision. The L-values are determined by the expected proportions of mice to be


of 39

paralysed with the reference vaccine. Laboratories should generate their own L-values following

the training and implementation procedure, and update them as they acquire data from further

tests. L-values should cover all tests conducted by the laboratory, but to detect variability

between operators, laboratories may wish to additionally calculate L-values for individual

operators.

The number of mice required for the TgmNVT was determined through the original collaborative

studies to ensure high probability of failing test vaccines with a specific level of neurovirulence.

The statistical analysis includes validity criteria such as a test for parallelism of dose response

curves (the “Vaccine x Dose interaction”). A step by step decision procedure is defined to check

the validity criteria, and come to a pass/fail decision. Provision exists for retests, either because

the test is statistically invalid, or because the LOR falls between the L1 and L2 values. The

decision procedure has been designed to reduce the need for retesting (and animal usage) in

situations where the test is statistically non-parallel, but a decision on the test vaccine can still be

made. For example, if the test vaccine is highly neurovirulent compared to the reference at both

doses, it can be considered a fail even if the dose-response lines are non-parallel.

A detailed description of the statistical model and the decision procedure is given in Appendix 4.

Analysis will require a suitable statistical software package for fitting the logistic regression

model, and collaboration with a qualified statistician may be advisable to ensure the correct

analysis and interpretation of test results.


of 39

Appendix 1: Randomization procedures for the Transgenic Mouse Neurovirulence Test

(TgmNVT)

Introduction

The recommended WHO procedure for testing vaccines in the transgenic mouse neurovirulence

test (TgmNVT) requires randomization of the mice to the doses, randomization of the cage

locations and randomization of the order of inoculation. A randomization process is provided

below for the following three cases which may be encountered during training and

implementation procedures as well as during batch release testing. Calculation spreadsheets to

perform the required randomization procedures are available from NIBSC on request.

• Case 1: one test vaccine and one reference vaccine, each at two doses, male and female

mice caged separately with 4 mice per cage, 8 mice per gender-dose group, inoculation on

one day. A total of 64 mice are used for this test.

• Case 2: one test vaccine and one reference vaccine, each at two doses, male and female

mice caged separately with 4 mice per cage, 16 mice per gender-dose group, inoculation

usually only on a single day. A total of 128 mice are used for this test.

• Case 3: two test vaccines and one reference vaccine, each at two doses, male and female

mice caged separately with 4 mice per cage, 16 mice per gender-dose group, inoculation

carried out over the course of two days. A total of 192 mice are used for this test.

General Principles of the Randomization

Full randomization is an important element of the TgmNVT and the random assignment of

animals to dose, cage location, and order of inoculation should be assured at the onset of the test.

However, the randomization procedure does not generate a complete randomization for each of

combination of factors and is actually a “balanced randomization.” These procedures aim to

provide:

• randomization of mice to doses,

• randomization of cage locations and

• randomization of order of inoculation.

The randomization procedure can be clarified by examining Case 1 below (Tables Q1, 2 and 3), a

64 animal test with 4 doses, i.e. 8 animals per sex/dose group.

When starting the randomization, the male mice are randomized first. There will eventually be 8

cages of male mice and so Column 1 of Table Q1 distributes the first 16 mice in the first 4 cages

– with one cage for each dose. The next 16 mice are distributed into the next 4 cages – again 1

cage per dose. Likewise with the females (Table Q2), four cages of females from the first 16 and

another 4 from the second 16.

When the male cages are randomized in the cage location (Table Q3), the first four cages are in

the first row and the second 4 in the second row. Likewise with the females, the first 4 cages in

Row 1 and the second 4 in Row 2.

The following outlines some of the features of the randomization:


of 39

1. Individual mice are randomly assigned to doses/vaccines, so that the experimental unit

is the mouse rather than the cage. This is preferred over randomizations of cages to

doses/vaccines because it is generally more efficient and it allows analysis to be carried out

on individual mouse observations instead of the more complex analysis based on cage

scores.

2. Doses, vaccines and gender are not confounded with each other or with any other

factors such as order of inoculation, cage location, or weight classifications. That is,

statistically valid estimates and tests can be carried out for dose effects, vaccine

differences, gender differences, Vaccine x Dose interactions, Vaccine x Gender

interactions and Dose x Gender since they are not inextricably mixed up with each other or

with other factors which may affect response.

3. A restricted randomization is applied to order of inoculation. This should provide

adequate protection against bias or variance changes which may be associated with order

effects. A complete randomization of inoculation order was viewed as impractical and

may induce errors in experimental procedures.

4. If shipment boxes have mice grouped according to gender, weight or age (i.e. factors

which are either known or suspected to affect response) the design retains this

homogeneity in assignments of mice to doses/vaccines (provided that the numbers of mice

within the homogeneous subgroups are consistent with the study design). That is, the

randomization is equivalent to randomly assigning doses/vaccines to mice within blocks or

strata defined by gender/weight/age groups.

5. Order of inoculation, weight/age groupings and location (rows) are at least partially

confounded. That is, the effects of these factors are inextricably mixed up so that it is

statistically impossible to differentiate the effects of one factor from the effects of the other

two factors. Some degree of confounding is necessary because of practical, logistic and

budgetary limitations and these three factors were considered to be of lower interest than

factors such as dose, vaccine and gender.

Case 1:

The randomization process for assigning mice to doses and for assigning cage locations is

provided in detail below. The randomization of the order of inoculation is accomplished by the

randomization of cage locations and the specified inoculation instructions. The instructions for

Case 1 are based on the randomizations provided in Tables Q1, Q2 and Q3. For each new

experiment the first step is to generate a new set of Tables Q1, Q2 and Q3 as follows:

• Randomly permute (i.e. randomly re-order) the assigned cages in each column of Table

Q1. The entries in each column of the new Table will be the same as in Table Q1 but the

order will be different.

• Randomly permute the assigned cages in each column of Table Q2.

• Randomly permute the cage designations in each row of Table Q3.

After generating the new tables follow the steps outlined in the sections “Randomization

Procedure” and “Inoculation Procedures” shown below.


of 39

Table Q1

Column 1 Column 2

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

1 Dm4 17 Am5

2 Bm2 18 Cm7

3 Cm3 19 Am5

4 Bm2 20 Dm8

5 Cm3 21 Dm8

6 Am1 22 Am5

7 Dm4 23 Dm8

8 Am1 24 Cm7

9 Cm3 25 Bm6

10 Bm2 26 Bm6

11 Dm4 27 Cm7

12 Am1 28 Am5

13 Dm4 29 Dm8

14 Am1 30 Bm6

15 Bm2 31 Cm7

16 Cm3 32 Bm6

Table Q2

Column 1 Column 2

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

33 Af9 49 Bf14

34 Df12 50 Df16

35 Bf10 51 Af13

36 Af9 52 Bf14

37 Df12 53 Af13

38 Af9 54 Bf14

39 Af9 55 Bf14

40 Cf11 56 Af13

41 Bf10 57 Df16

42 Df12 58 Af13

43 Cf11 59 Cf15

44 Cf11 60 Cf15

45 Bf10 61 Df16

46 Cf11 62 Cf15

47 Df12 63 Cf15

48 Bf10 64 Df16

Table Q3

Row 1 Bm2 Am1 Df12 Af9 Cm3 Bf10 Dm4 Cf11

Row 2 Bf14 Cm7 Am5 Af13 Cf15 Df16 Bm6 Dm8

Randomization Procedure

Step 1. The mice will be housed 4/cage with males and females in separate cages. The cages

will be labeled as follows:

Male Mice Female Mice

Am1 Bm2 Cm3 Dm4 Af9 Bf10 Cf11 Df12


That is, cages are numbered from 1 to 16, with numbers 1-8 for males (coded ‘m’) and 9-16 for

females (coded ‘f’); the letters A, B, C and D indicate the coded (blinded) dose/vaccine which

will be inoculated into the mice within the cage.


of 39

Step 2. It is assumed that mice are delivered to the lab in boxes of 18-20 mice which are of the

same gender and similar weight/age. Arrange the shipment boxes in the following order: start

with the boxes containing heavier males, followed by boxes containing lighter males, followed by

boxes containing heavier females and then lighter females. In many instances there are no

differences in the age/weight of animals in each of the shipping boxes, in this instance just

arrange the boxes in a random order i.e. as they come to hand, with males first and females

second.

Step 3. Arrange cages Am1 Bm2 Cm3 and Dm4 in a convenient manner to receive mice. Take

16 mice from the first shipment box and place them into these cages in the sequence provided in

the first column of Table Q1. That is, the first mouse that comes to hand is placed in cage Dm4;

the second mouse is placed in cage Bm2, the third mouse in cage Cm3, etc., the 16th mouse in

cage Cm3. Cages Am1 Bm2 Cm3 and Dm4 now have 4 mice/cage.

Step 4. Arrange cages Am5 Bm6 Cm7 and Dm8 to receive the next 16 mice. As mice come to

hand place them in these cages in the sequence provided in the second column of New Table Q1.

Step 5. Repeat Steps 3-4 for the female mice using cages Af9 Bf10 ... Df16 and the caging

sequence provided in Table Q2.

Step 6. Place the cages in racks in the positions identified in Table Q3.

Inoculation Procedures

Step 7. Prior to inoculation the mice should be weighed. Gender, weight/age, cage identification,

location, dose/vaccine, and outcome (paralysis) must be recorded for each mouse. Records must

also be maintained of the date mice arrive at the lab, date of caging and inoculation, results of

back-titration, any losses and reasons for the losses (e.g. accidental deaths) and any deviation

from protocol (e.g. individual housing of overly aggressive mice).

Step 8. Randomize the doses to be used in the inoculation and assign them the letters A, B, C and

D. To minimize bias in assessing behavioral clues during inoculation and scoring, neither the

inoculator(s) nor clinical scorer(s) should be made aware of which vaccine or dose corresponds to

which letter until data collection is complete.

Step 9. Starting with the cage at the left end of Row 1 (see Table Q3), inoculate the 4 mice in this

cage with the dose/vaccine specified on the cage. Then inoculate the mice in the next cage of

Row 1, etc until Row 1 is completed. Proceed to Row 2 starting at the left and complete

inoculations for Row 2. All inoculations are to be performed by one person on a single day.

Case 2:

This case relates to a test of vaccines versus reference with 16 animals per gender/dose group i.e.

a total of 128 mice. The instructions for Case 2 are based on the randomizations provided in

Tables D1, D2 and D3.

For each new experiment the first step is to generate a new set of Tables D1, D2 and D3 as

follows:


of 39

Table D1

Column 1 Column 2 Column 3 Column 4

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

1 Dm4 17 Am5 33 Am9 49 Bm14

2 Bm2 18 Cm7 34 Dm12 50 Dm16

3 Cm3 19 Am5 35 Bm10 51 Am13

4 Bm2 20 Dm8 36 Am9 52 Bm14

5 Cm3 21 Dm8 37 Dm12 53 Am13

6 Am1 22 Am5 38 Am9 54 Bm14

7 Dm4 23 Dm8 39 Am9 55 Bm14

8 Am1 24 Cm7 40 Cm11 56 Am13

9 Cm3 25 Bm6 41 Bm10 57 Dm16

10 Bm2 26 Bm6 42 Dm12 58 Am13

11 Dm4 27 Cm7 43 Cm11 59 Cm15

12 Am1 28 Am5 44 Cm11 60 Cm15

13 Dm4 29 Dm8 45 Bm10 61 Dm16

14 Am1 30 Bm6 46 Cm11 62 Cm15

15 Bm2 31 Cm7 47 Dm12 63 Cm15

16 Cm3 32 Bm6 48 Bm10 64 Dm16

Table D2


Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

65 Af17 81 Af21 97 Af25 113 Af29

66 Cf19 82 Bf22 98 Cf27 114 Df32

67 Bf18 83 Bf22 99 Af25 115 Bf30

68 Af17 84 Df24 100 Df28 116 Af29

69 Df20 85 Cf23 101 Df28 117 Cf31

70 Bf18 86 Bf22 102 Bf26 118 Bf30

71 Cf19 87 Df24 103 Bf26 119 Df32

72 Df20 88 Af21 104 Cf27 120 Cf31

73 Bf18 89 Cf23 105 Af25 121 Cf31

74 Bf18 90 Bf22 106 Bf26 122 Df32

75 Af17 91 Af21 107 Df28 123 Bf30

76 Df20 92 Cf23 108 Bf26 124 Af29

77 Cf19 93 Af21 109 Cf27 125 Bf30

78 Cf19 94 Cf23 110 Af25 126 Df32

79 Df20 95 Df24 111 Df28 127 Cf31

80 Af17 96 Df24 112 Cf27 128 Af29


of 39

Table D3

Row 1 Bm2 Am1 Df20 Af17 Cm3 Bf18 Dm4 Cf19

Row 2 Bf22 Cm7 Am5 Af21 Cf23 Df24 Bm6 Dm8

Row 3 Cm11 Af25 Bf26 Df28 Am9 Dm12 Bm10 Cf27

Row 4 Am13 Bm14 Bf30 Af29 Dm16 Cm15 Cf31 Df32


D1. The entries in each column of the new Table will be the same as in Table D1 but the


• Randomly permute the assigned cages in each column of Table D2.

• Randomly permute the cage designations in each row of Table D3.




Randomization for an experiment with a test vaccine and a reference vaccine, each at two doses

with 16 mice per gender/dose/vaccine group.

Step 1. The mice will be housed 4/cage with males and females in separate cages. The cages

will be labeled as follows:

Male Mice Female Mice






females (coded ‘f’); the letters A, B, C and D indicate the dose/vaccine which will be inoculated

into the mice within the cage. i.e. vaccine and reference, 2 doses of each.




boxes containing heavier females and then lighter females. If there is no age/size difference

between mice then randomly arrange male and female boxes.

Step 3. Arrange cages Am1 Bm2 Cm3 and Dm4 in a convenient manner to receive mice. Take

16 mice from the first shipment box and place them into these cages in the sequence provided in

the first column of Table D1. That is, the first mouse that comes to hand is placed in cage Dm4;

the second mouse is placed in cage Bm2, the third mouse in cage Cm3, etc., the 16th mouse in

cage Cm3. Cages Am1, Bm2, Cm3, and Dm4 now have 4 mice/cage.


of 39

Step 4. Arrange cages Am5 Bm6 Cm7 and Dm8 to receive the next 16 mice. As mice come to

hand place them in these cages in the sequence provided in the second column of Table D1.

Step 5. Repeat the Step 4 with cages Am9 Bm10 Cm11 and Dm12 using the sequence provided

in the third column of Table D1.

Step 6. Repeat the Step 4 with cages Am13 Bm14 Cm15 and Dm16 using the sequence provided

in the fourth column of Table D1.

Step 7. Repeat Steps 3-6 for the female mice using cages Af17 Bf18 ... Df32 and the caging

sequence provided in New Table D2.

Step 8. Place the cages in racks in the positions identified in Table D3.







Step 10. Randomize the doses to be used in the inoculation and assign them the letters A, B, C

and D. To minimize bias in assessing behavioral clues during inoculation and scoring, neither the

inoculator(s) nor clinical scorer(s) should be aware of which vaccine or dose corresponds to


Step 11. Starting with the cage at the left end of Row 1 (see Table D3), inoculate the 4 mice in

this cage with the dose/vaccine specified on the cage. Then inoculate the mice in the next cage of

Row 1, etc until Row 1 is completed. Proceed to Row 2 starting at the left and complete

inoculations for Row 2. All inoculations are to be performed by one person on a single day.

Case 3: This case is based on a test for 2 vaccines versus the reference with 16 mice per dose/gender

group. A total of 192 mice are used for this study. The instructions for Case 3 are based on the

randomizations provided in Tables F1, F2 and F3. For each new experiment the first step is to

generate a new set of Tables F1, F2 and F3 as follows:


F1. The entries in each column of the new Table will be the same as in Table F1 but the


• Randomly permute the assigned cages in each column of Table F2.

• Randomly permute the cage designations in Rows 1 and 2 of Table F3 and then randomly

permute the cage designations of the Rows 3 and 4, Rows 5 and 6 and Rows 7 and 8 of

Table F3




of 39


Randomization for an Experiment with 2 test vaccines and a reference vaccine, each at two doses

with 16 mice per gender/dose group i.e. a total of 192 mice.

Step 1. The mice will be housed 4/cage with males and females in separate cages. The cages will

be labeled as follows:

Male mice

Am1 Bm2 Cm3 Dm4 Em5 Fm6




Female mice

Af25 Bf26 Cf27 Df28 Ef29 Ff30




Table F1


Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

1 Bm2 25 Dm10 49 Em17 73 Dm22

2 Fm6 26 Em11 50 Bm14 74 Em23

3 Dm4 27 Cm9 51 Fm18 75 Bm20

4 Am1 28 Em11 52 Cm15 76 Fm24

5 Fm6 29 Am7 53 Bm14 77 Am19

6 Cm3 30 Am7 54 Fm18 78 Em23

7 Em5 31 Fm12 55 Cm15 79 Cm21

8 Bm2 32 Am7 56 Am13 80 Bm20

9 Cm3 33 Dm10 57 Am13 81 Am19

10 Am1 34 Bm8 58 Am13 82 Bm20

11 Am1 35 Am7 59 Am13 83 Dm22

12 Fm6 36 Dm10 60 Dm16 84 Cm21

13 Em5 37 Bm8 61 Em17 85 Dm22

14 Em5 38 Fm12 62 Cm15 86 Fm24

15 Am1 39 Bm8 63 Fm18 87 Bm20

16 Bm2 40 Dm10 64 Em17 88 Fm24

17 Bm2 41 Cm9 65 Fm18 89 Cm21

18 Dm4 42 Em11 66 Dm16 90 Am19

19 Dm4 43 Fm12 67 Bm14 91 Dm22

20 Em5 44 Em11 68 Dm16 92 Em23

21 Cm3 45 Bm8 69 Dm16 93 Fm24

22 Cm3 46 Fm12 70 Em17 94 Cm21

23 Fm6 47 Cm9 71 Bm14 95 Em23

24 Dm4 48 Cm9 72 Cm15 96 Am19


of 39


females (coded ‘f’); the letters A, B, C, D, E and F indicate the dose/vaccine which will be

inoculated into the mice within the cage.




boxes containing heavier females and then lighter females. If there is no age/size difference

between mice then randomly arrange male and female boxes.

Step 3. Arrange cages Am1 Bm2 Cm3 Dm4 Em5 and Fm6 in a convenient manner to receive

mice. Take 24 mice from the first 2 shipment boxes and place them into these cages in the

sequence provided in the first column of Table F1. That is, the first mouse that comes to hand is

placed in cage Bm2; the second mouse is placed in cage Fm6, the third mouse in cage Dm4, etc.,

the 24th mouse in cage Dm4. Cages Am1, Bm2, Cm3, Dm4, Em5, and Fm6 now have 4

mice/cage.

Table F2


Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

Mouse

Handling

Order

Assigned

Cage

97 Cf27 121 Cf33 145 Bf38 169 Bf44

98 Bf26 122 Bf32 146 Ef41 170 Df46

99 Af25 123 Af31 147 Ff42 171 Af43

100 Df28 124 Df34 148 Df40 172 Af43

101 Cf27 125 Cf33 149 Bf38 173 Df46

102 Ff30 126 Ff36 150 Cf39 174 Ef47

103 Ef29 127 Ef35 151 Ff42 175 Ff48

104 Df28 128 Df34 152 Cf39 176 Bf44

105 Cf27 129 Cf33 153 Cf39 177 Ff48

106 Ff30 130 Ff36 154 Df40 178 Ff48

107 Cf27 131 Cf33 155 Af37 179 Ef47

108 Df28 132 Df34 156 Bf38 180 Ef47

109 Bf26 133 Bf32 157 Df40 181 Ef47

110 Ff30 134 Ff36 158 Df40 182 Cf45

111 Df28 135 Df34 159 Bf38 183 Cf45

112 Ef29 136 Ef35 160 Ef41 184 Df46

113 Ef29 137 Ef35 161 Af37 185 Bf44

114 Bf26 138 Bf32 162 Af37 186 Ff48

115 Af25 139 Af31 163 Af37 187 Cf45

116 Ef29 140 Ef35 164 Ff42 188 Af43

117 Af25 141 Af31 165 Ef41 189 Af43

118 Bf26 142 Bf32 166 Ef41 190 Bf44

119 Af25 143 Af31 167 Cf39 191 Cf45

120 Ff30 144 Ff36 168 Ff42 192 Df46


of 39

Step 4. Arrange cages Am7 Bm8 Cm9 Dm10 Em11 and Fm12 to receive the next 24 mice. As

mice come to hand place them in these cages in the sequence provided in the second column of

Table F1.

Step 5. Repeat Step 4 with cages Am13 Bm14 Cm15 Dm16 Em17 and Fm18 using the sequence

provided in the third column of New Table D4.

Step 6. Repeat the Step 4 with cages Am19, Bm20, Cm21, Dm22, Em23, and Fm24 using the

sequence provided in the fourth column of New Table D4.

Step 7. Repeat Steps 3-6 for the female mice using cages Af25, Bf26, Cf27 .... Ff48 using the

caging sequence provided in Table F2.

Step 8. Place the cages in racks in the positions identified in Table F3.

Table F3

Row 1 Df28 Bf26 Ff30 Dm4 Am1 Bm2

Row 2 Em5 Cf27 Cm3 Af25 Fm6 Ef29

Row 3 Bm8 Df34 Cf33 Af31 Fm12 Bf32

Row 4 Ff36 Ef35 Em11 Dm10 Cm9 Am7

Row 5 Ef41 Bm14 Cm15 Cf39 Bf38 Fm18

Row 6 Af37 Em17 Df40 Am13 Dm16 Ff42

Row 7 Df46 Bf44 Fm24 Af43 Ef47 Em23

Row 8 Cm21 Cf45 Ff48 Bm20 Am19 Dm22







Randomize the doses to be used in the inoculation and assign them the letters A, B, C, D, E and

F. To minimize bias in assessing behavioral clues during inoculation and scoring, neither the

inoculator(s) nor clinical scorer(s) should be aware of which vaccine or dose corresponds to


Step 11. Starting with the cage at the left end of Row 1 (see Table D3), inoculate the 4 mice in

this cage with the dose/vaccine specified for the cage. Then inoculate the mice in the next cage

of Row 1, etc until Row 1 is completed. Proceed to Row 2 starting at the left and complete

inoculations for Row 2.

Step 12. Inoculation for this study can be completed in a single day; however, it is generally

found that it is more appropriate to split the inoculations across 2 days. If this is to be done then:


of 39

• Rows 1-4 should be inoculated on Day 1 and Rows 5-8 on Day 2.

• A separate vial of each of the six samples should be used on each of the days that

inoculations are carried out.

• Either one inoculator should perform all injections over two days, or if two inoculators are

needed, all injections on a day should be performed by the same inoculator.


of 39

Appendix 2: Recommended training program for the Transgenic Mouse Neurovirulence

Test (TgmNVT) and for the training of new inoculators and maintenance of competence of

qualified inoculators

Introduction

The initial training program for an inoculator/organization consists of three steps. Each

consecutive step should be made only after the previous step has been accomplished.

The Training Steps

Step 1. Obtain training in the technique of intraspinal inoculation of mice at a WHO

recognized training laboratory. If the organization has not obtained training in the

clinical scoring of the TgmNVT they should also be trained in this area during Step 1.

Step 2. Inoculate conventional mice intraspinally with 5 µl of 10% India ink, dissect out the

spinal cord and examine for the presence of the ink in the correct region of the cord.

Step 3. Conduct two independent tests in TgPVR21 mice for each of Serotypes 1, 2 and 3.

An organization need only complete the training tests for each serotype once.

Training of new inoculators will involve a reduced set of tests designed to

demonstrate the competence of the inoculator to perform the procedure rather than

demonstrate the ability of the organization to undertake the test in full.

Preliminary Training- Step 1

Training for Inoculators

WHO can provide information on laboratories that will be able to assist in the preliminary

training phase for an organization undertaking the training and implementation for the first time.

Training in clinical scoring

Each of the three serotypes can manifest subtly different clinical symptoms and it is therefore

important that an organization undertaking training in clinical scoring attempts to observe and

score tests of all three serotypes during the training phase. The suggested procedure for training

in the clinical scoring is:

1) An observation/explanation phase with a qualified operator (usually 1 or 2 tests),

2) Assisted scoring sessions where the qualified operator and the trainee score tests together

(usually 1 or 2 tests),

3) At this stage the trainee and the trainer score the same animals independently in a mock test

situation (i.e. a full scoring day). The scoring should include at least forty animals with clinical

signs of polio infection. Timing between the scoring by the trainer and the trainee should be

minimised to reduce the possibility that the disease will progress between scoring sessions. The

trainer should always score the animals before the trainee. At the end of scoring the score sheets

are compared. Any differences are confirmed in a second reading by the trainer and where

appropriate the corrected scores are used to calculate the concordance level. To be considered

competent the result of the trainee’s scoring should be evaluated in comparison to the trainer’s for

the number of scoring difference (D) over the total number of animals (N) (D/N) and the number

of scoring differences involving the change from weak to partial paralysis 1 (Dw-pp) over the

number of animals with clinical signs (CS) (Dw-pp/CS). For D/N and Dw-pp/CS the trainee

should not differ more than 5% from the trainer’s score. If the trainee does not meet the

established criteria they should begin again with at least step 2 of the procedure. If the exercise

must be repeated again due to an unsatisfactory score, only the most recent data is considered in

the next calculation of the scoring difference.


of 39

A trainee should also be able to adequately evaluate the inoculation step (including the evaluation

of the documentation related to randomisation of mice, tracking of inocula and mice/cages etc.)

and perform the calculations of the results. Appropriate training data sets for calculation and

example documentation on inoculation should be used to ensure that the correct results are

obtained by the trainee. (NIBSC is willing to provide the appropriate training data sets if

requested by the OMCL). NIBSC is willing to act as the training centre for any of the three stages

noted above upon arrangement with the interested OMCL and could if needed provide a letter

indicating that the step(s) had been completed successfully.

In preparation for the hands on training, in a preliminary phase, the trainee should watch the

training DVD on clinical evaluation of oral polio vaccine testing in transgenic mice prepared by

and available from NIBSC.

It is recommended that the above training all be completed within a 6 month period and that

adequate records be kept of all training sessions.

Within an organization regular parallel scoring sessions should be undertaken as described above

to ensure that members of staff are consistent and agree on the clinical criteria to be applied. It is

also recommended that whenever possible organizations conduct training in clinical scoring with

other organizations to avoid shifts in interpretation and application of the criteria that can occur

over time. In the case of a long period of scoring inactivity (greater than 6 months) measures

should be taken to re-qualify observation skills before a test scoring session is undertaken.

Dye Inoculation –Step 2

Mice are inoculated using the same procedure as the standard test, but instead of vaccine, a 10%

India ink solution is injected to check placement of the needle and assess the extent of diffusion

of injected material. Mice are euthanized immediately after inoculation and the spinal column

removed and fixed in 10% formalin for 2 days. Alternatively the spine can be frozen and several

blocks of the frozen spinal column prepared using a sharp blade. The cord is first sectioned

transversely between the last thoracic and the first lumbar vertebrae. Then two blocks are

prepared from the low thoracic region and three blocks from the lumbar region, each about 3 to 5

mm long, and the blocks examined. Each test should consist of at least 30 animals.

Animals have been correctly inoculated if there is an India ink deposit in the anterior gray matter

and there is no diffusion of the India ink solution outside of the area immediately surrounding the

point of inoculation. An inoculator is considered competent to pass to the next stage of the

training if ≥ 90% of animals are inoculated correctly in each of 3 consecutive tests.∗

The Training Tests – Step 3

Serotype 1

Training for Type 1 involves two separate tests of a Type 1 preparation alongside the WHO/I

reference preparation. The dilutions producing the required doses for both vaccine and reference

will be given in the documentation provided with the samples.

∗ There is no general rule about how many mice and how long it may take for any given inoculator to become

competent i.e. to achieve >90% accuracy in 3 consecutive tests. However, with regular training sessions (at 2-4 week

intervals) and adequate numbers of mice in each session (30 or more) experience indicates that the desired accuracy

over three consecutive tests may require up to 20 tests.


of 39

Dilutions are made as described in the relevant documentation and the design and randomization

procedure is as provided in Appendix 1, Case 2 with 16 mice/dose-gender. Each study must

satisfy validity criteria c1, c2, c3:

c1 The WHO/I reference preparation paralysis rates at doses 1.75 and 2.75 must be between

0.05 and 0.95;

c2 The WHO/I reference vaccine must have a statistically significant (p<0.05, one-tailed

test) increase in paralysis rates from the 1.75 to the 2.75 dose;

c3 The test for Vaccine x Dose interaction must have a p-value >0.05 (i.e. no evidence of a

Vaccine x Dose interaction).

Serotype 2

Training for type 2 involves two separate tests of a single vaccine and the WHO/II reference

preparation. The dilutions producing the required doses for both vaccine and reference will be

given in the documentation provided with the samples.

Dilutions are made as described in the relevant documentation and the design and randomization

procedure is as provided in Appendix 1, Case 2 with 16 mice/dose-gender. Each study must

satisfy validity criteria c1, c2, c3:

c1 The WHO/II reference vaccine paralysis rates at doses 5.0 and 6.0 must be between 0.05

and 0.95;

c2 The WHO/II reference vaccine must have a statistically significant (p<0.05, one-tailed




Serotype 3

Training for type 3 requires inoculation of two different vaccines in the two tests. The first test

uses a vaccine with a markedly elevated amount of a known neurovirulent mutation at position

472 of the viral genome (472-C revertants). This sample is tested alongside the WHO/III

reference preparation. The dilutions producing the required doses for both vaccine and reference

will be given in the documentation provided with the samples.

The design and randomization procedure is as provided in Appendix 1, Case 1 with a reduced

number of mice (8 mice/dose-gender) as this vaccine should be a clear fail in the test. The study

must satisfy the following validity criteria:

c1 The WHO/III reference vaccine paralysis rates at doses 3.5 and 4.5 must be between 0.05

and 0.95; c2 The WHO/III reference vaccine must have a statistically significant (p<0.05, one-tailed


c3 The decision model applied to the paralysis data must result in rejection of the test

vaccine.

If this test is satisfactory a second test is carried out with a vaccine with a moderately elevated

amount of 472-C revertants and the WHO/III reference vaccine. The design and randomization


of 39

procedure provided in Appendix 1, Case 2 with 16 mice/dose-gender should be followed. The

study must satisfy validity criteria c1, c2 and c3 in Step 4 and the following additional criterion:



Return of Data

Upon completion of a test the following data should be provided to the WHO-affiliated reference

laboratory. The NCL may also require submission of these data at this time or at a later stage.

The following data should be included:

• Randomization scheme used showing the allocation of mice of each sex by cage number, the

allocation of vaccine/dose sample by cage, the order in which cages were inoculated and any

additional data e.g. dates on which each cage were inoculated;

• Daily clinical record sheets for all mice;

• Summary of the results from the test. These should be provided in a spreadsheet in which

following are presented for each mouse:

For the initial training phase for the TgmNVT the following data are not required, but may be

submitted:

• Titration data for the dilutions and the titration reference prior to inoculation into mice.

• Titration data for the residual inocula.

• Decisions on validity of the test and the pass/fail outcome for each sample.

Training for New Inoculators

For organizations that have completed the implementation process for at least one virus serotype

the training process for a new inoculator is:

Step 1. Obtain training in the technique of intraspinal inoculation of mice with a qualified

inoculator within the organization.

Step 2. Inoculate conventional mice intraspinally with 5 µl of 10% India ink, dissect out the

spinal cord and examine for the presence of the ink in the correct region of the cord.

The success criteria at this step are the same as outlined above, i.e. 3 consecutive tests

of 30 or more mice with an accuracy of 90% or greater.

Step 3. Training tests: only three tests with viruses from the training or implementation panels

need to be completed, with the expected result being obtained in each of the three tests.

The schedule for tests to be undertaken should be agreed with the WHO affiliated

reference laboratory.

Mouse

Numbe

r

Sex

(M/F)

Mouse

weight

Cage Sample

Inoculated

Dose Inoculator Date

Inoculated

Clinical

stage

(max)

Paralysis

score

(0/1)

1 M 28.18 Dm4 WHO/III 4.5 FB 29/2/04 paresis 1

2 …. …


of 39

Step 4. The WHO affiliated reference laboratory should provide details of the test samples to

the NRA and the results of the tests should be provided by the testing organization to

the NRA.

Maintenance of Competence

Following qualification, inoculators need to continue inoculating to retain the competence. If a

qualified inoculator has not completed a test within the previous 3 months they should validate

their technique by completing a series of intraspinal inoculations with India ink. Not less than 30

mice should be inoculated and an accuracy of 90% or greater must be achieved. Failure to

achieve 90% accuracy will require that they carry out the India ink inoculation phase of this

training procedure until they have reached an accuracy of 90% or greater on three consecutive

tests. After achieving this accuracy the inoculator can then resume TgmNVT inoculations.

Each inoculator shall perform a minimum of one India ink inoculation per year. Records of all

tests carried out for maintenance of competence should be retained and made available for the

NCL if requested. DVDs with recordings of mice at the different stages of clinical signs of paralysis will be used as

a teaching and monitoring aid.Two types of DVD are available:

1) Training and refreshing observation skills: For a novice scorer, to familiarise them with

the different stages and provide harmonised examples and for experienced scorers to refresh

their visual memory, especially when there has been a significant time lapse since the scorer

has last evaluated a test. This style of DVD clearly identifies examples and provides

commentary.

2) PTS-like exercise: In order to evaluate the maintenance of competence DVDs will be

prepared with a compilation of images of mice at different stages. The stages of the mice will

not be identified. The reader will complete a score sheet and submit it to the organiser for

evaluation. If the score does not fall within a given limit of difference (to be determined based

on NIBSC data score sheets from different operators) the scorer would be required to go for

additional training. NIBSC has committed to preparing the DVDs for both purposes. PTS-like

studies will be run at least once per year.


of 39

Appendix 3: Recommended implementation process for the Transgenic Mouse

Neurovirulence Test (TgmNVT)

Introduction

Following successful completion of the “Recommended Training Program for the Transgenic

Mouse Neurovirulence Test (TgmNVT)” (see Appendix 2), WHO recommends that laboratories

complete the standard implementation process to validate its readiness to perform the TgmNVT

on commercial vaccine batches. For this purpose the laboratory must conduct three qualifying

experiments per serotype with a coded proficiency panel supplied by a WHO nominated

reference laboratory. Each panel consists of six samples and two samples are tested concurrently

against the homotypic WHO reference virus in each of three tests. The appropriate design and

randomization protocols for these tests are provided in Appendix 1, Case 3.

Provision of Samples

Once training is successfully completed the standard implementation panel may be requested

from WHO affiliated reference laboratory. For each serotype the panel consists of six coded

samples that have been previously tested in both the monkey and mouse NVT. The samples are

selected to provide a stringent evaluation of the performance of the mouse test. No information is

provided on the titre or expected result for each sample. Laboratories are requested to test all six

samples then to send their raw data and pass/fail decisions to the WHO affiliated reference

laboratory for evaluation.

Titration of Samples

In parallel with the training process in mice, laboratories should document the precision and

evaluate the sensitivity of their viral titration procedures for each serotype they subsequently test

in the TgmNVT. Reference preparations for viral titrations can be provided by a WHO affiliated

reference laboratory for this purpose. Because of the sensitivity of the mouse neurovirulence test

to virus dose, laboratories should achieve for each serotype a precision of ± 0.3 log10 CCID50/ml

or better for the confidence limits of the mean titre. The observed titres should also be compared

with the assigned titres of the titration reference preparations.

Calibration of an in-house titration reference

It is recommended that laboratories calibrate their own in-house reference against the available

titration references for use in assigning titres for the TgmNVT.

Tests in mice

The implementation process requires that each sample be tested once in mice. Each test should

include the WHO reference preparation plus two samples and will require 192 mice. “Instructions

for Use” will accompany each panel and will explain which samples are to be included in each of

the three tests.

A minimum of three tests is required to complete the implementation process for each serotype.

As part of the process organizations are requested to make an assessment of the validity of each

test and, for valid tests, whether each sample passes or fails. When three valid tests are

completed, raw data from all tests performed in the qualification process (valid and invalid)

should be forwarded to the WHO affiliated reference center coordinating the implementation

process for further evaluation. The data and format to be provided will be given in the

“Instructions for Use” that accompany each of the panels.


of 39

Return of Data

Upon completion of the tests for each implementation panel the following data should be

provided:

• Titration data for each of the samples, the WHO(SO+2) standard, and the titration reference.

Data should be provided for each titration undertaken;

• Titration data for the dilutions and the titration reference measured prior to inoculation into

mice. Data should be provided for each titration undertaken;

• Titration data for the residual inocula measured after inoculation into mice. Data should be

provided for each titration undertaken;

• Randomization scheme used showing the allocation of mice of each sex by cage number, the

allocation of vaccine/dose sample by cage, the order in which cages were inoculated and any

additional data, e.g., dates on which each cage were inoculated;

• Daily clinical scoring records for all mice;

• Summary of the results from the test. These should be provided in a spreadsheet in which

data are presented for each mouse, as shown below. Additional data such as unexpected

mortality should also be recorded;

• Decisions on validity of the test and the pass/fail outcome for each sample;

• Recalculated L values and the numbers used to recalculate these values.

Evaluation

The evaluation will initially involve re-assessment of whether each test meets validity criteria

specified in the SOP. These are:

• that the doses of virus inoculated into mice should be within ± 0.3 log10 CCID50 of the

target dose;

• the paralysis proportions of the WHO reference are within limits;

• there is a significant increase in paralysis for the WHO reference from the low to the high

dose;

• that there is no evidence for a Vaccine x Dose interaction. The evaluation will also

compare the observed outcome (pass or fail) with the expected outcome.

In addition the data will be reviewed to determine that there are no Vaccine x Gender and Dose x

Gender effects, that the estimated log odds ratios from the tests of the duplicate vaccines are not

significantly different from each other, and that the vaccines are successfully ranked according to

their expected neurovirulence. Any inconsistencies identified in these additional evaluations will

indicate a need for careful review of procedures rather than invalidate the performance of the

laboratory.

Criteria to be used to determine if a laboratory successfully completes the implementation

process

Satisfactory completion of the implementation process requires that an organization completes a

valid test for all six samples from a total of not more than five tests and obtains the expected

results for all six samples. Failure to meet these criteria will indicate a need for retraining. The

extent of the retraining shall be determined by the NRA in collaboration with the WHO affiliated

reference laboratory.

Introduction of the mouse test for batch release

A laboratory that has successfully completed the implementation process should use the

accumulated data from 5 valid tests (3 in this implementation process plus 2 from the training

assays) to determine their own L1 and L2 values and use these for batch release purposes (see also


of 39

Appendix 4). The accumulated data may be used to support a regulatory application to perform

the mouse TgmNVT.


of 39

Appendix 4: Statistical analysis and decision model for testing Serotype 1, 2 or 3 OPV

vaccines in transgenic mice

Introduction

Logit transformations of Paralysis Proportions are represented as linear functions of vaccine, dose

and study design parameters (e.g. lab, replicate within lab, inoculation day, gender). Parameter

estimates and significance tests are based on Maximum Likelihood procedures and log likelihood

chi-squared tests. To test an interaction, a chi-squared test is applied to the difference in log

likelihoods between the main effects model with and without the interaction term. Main effects

are tested in a similar manner by applying a chi-squared test to the difference in log likelihoods

between the main effects model with and without the main effect under consideration. The main

effects models are parameterized so that the estimates of the vaccine effects and dose effects are

represented as the log odds ratios. That is, the relative neurovirulence of a test vaccine and the

WHO (SO+2) reference is estimated by the log odds ratio. Estimates and standard errors derived

from Maximum Likelihood analyses of the main effects models can be used to calculate 95%

confidence intervals. Exponential transformation of the estimates and confidence limits provide

estimated odds ratios and 95% fiducial limits for the odds ratio. The odds ratio indicates how

much more likely it is for paralysis to occur in the test vaccine than in the reference vaccine.

The recommended methodology for comparing a test vaccine against the WHO reference

requires:

• the test vaccine be tested concurrently with the WHO reference vaccine in a

randomized experiment; and that

• both vaccines are to be tested at 2 doses, a high and low dose, as shown below:

High Dose Low Dose

Type 1 2.75 log10 CCID50/5µl 1.75 log10 CCID50/5µl

Type 2 6.0 log10CCID50/5µl 5.0 log10 CCID50/5µl

Type 3 4.5 log10 CCID50/5µl 3.5 log10 CCID50/5µl

• each dose is to be intraspinally inoculated into 16 male and 16 female mice;

• the doses inoculated are within the specified ranges, as defined below, and meet other

relevant criteria;

• randomization of mice to doses, cage locations and inoculation order are carried out as

described in Appendix 1.

The following sections provide some background to the statistical model, a decision model for

the acceptance of the inoculum doses used in the test, specifications and examples for the

statistical analysis of the paralysis data and the application of the pass/fail decision model

Analysis

If the experiment involves testing with more than one test vaccine, individual analyses of

paralysis data are carried out comparing each test vaccine with the concurrent WHO(SO+2)

Reference.

The analysis is also based on the assumption that the doses used in the test are the target doses

specified. To confirm this, a decision model is applied to ascertain that, within accepted levels of


of 39

variation, the actual titres do not differ from the target doses.

Decision Model – Paralysis Data.

STEP 1. Check the paralysis proportions for the reference. If the combined (male and

female) proportion at the high dose > 0.95 or the combined proportion at the low

dose < 0.05 the reference data are regard as questionable. The experiment must be

repeated (GO TO STEP 6).

STEP 2. Check estimability. If the Maximum Likelihood procedure fails to converge then

apply the pass/fail criteria at each dose. The test vaccine is accepted if the vaccine

passes at both doses (i.e. log odds ratio < L1). The test vaccine fails if the vaccine

fails at both doses (i.e. log odds ratio > L2). All other outcomes will require a re-

test (i.e. GO TO STEP 6).

STEP 3. Apply the Maximum Likelihood procedure to the logistic regression model to test

for Vaccine x Dose Interaction. If the interaction is significant then apply the

pass/fail criteria at each dose. The Test Vaccine is accepted if the vaccine passes

at both doses (i.e. logs odds ratio < L1 at each dose). The Test Vaccine fails if the

log odds ratio > L2 at both doses. All other outcomes will require a re-test (i.e. go

to STEP 6).

If the observed paralysis rates for the test vaccine doses are either at 0.0 or 1.0,

then the Maximum Likelihood procedure may not provided a valid test of

interaction. Provided that the validity criteria for the reference are acceptable, the

following decision process can be applied:

• If both doses of the test vaccine produce legitimate 0.0 paralysis rates then

the vaccine is accepted.

• If a 0.0 paralysis rate only occurs for the low dose and the high dose

paralysis rate is lower than that of the corresponding reference then the test

vaccine is accepted, otherwise proceed to STEP 5.

• If the test vaccine has paralysis rates of 1.0 at both doses, the test vaccine

fails.

• If the test vaccine has a paralysis rate of 1.0 only at the high dose, then the

decision process is applied to the combined results for both doses and also

to the log odds ratio for the low dose. The vaccine is required to pass the

decision criteria for both the combined estimate of the log odds ratio and

the estimate at the low dose.

STEP 4. Check for a significant (p<0.05, one-tailed test) Dose effect. If Dose effect and

Vaccine effect are not significant, repeat the experiment (i.e. GO TO STEP 6). If

the Dose effect is not significant but the Vaccine effect is significant, then apply

the criteria in STEP 5.


of 39

STEP 5. Calculate the Log Odds Ratio.

If Log Odds Ratio <= L1 Vaccine passes

If L1 < Log Odds Ratio < L2 GO TO STEP 6

If Log Odds Ratio => L2 Vaccine fails

As laboratories will not have established their own values for L1 and L2 the values initially set

from the collaborative studies are given below. These are the values that should be used for the

assessment of the training and implementation panels used in the procedures outlined in

Appendix 3.

L1 L2

Type 1 0.734 1.037

Type 2 0.665 0.940

Type 3 0.645 0.913

STEP 6. If a pass/fail decision is not reached in STEPS 1-5 and a repeat experiment is

required, the decision process will be applied to either the pooled data from the

two experiments or the data from the second experiment alone. The analysis

depends on the circumstances which led to the retest:

• If the retest was initiated due to a technical problem in the first experiment

or due to a lack of validity of the reference profile (e.g. failed Step 1 or

Step 4) then STEPS 1-5 must be repeated using the tests and estimates

from the second experiment alone.

• If the retest was initiated because the Log Odds Ratio is between L1 and

L2 or because of a problem with the test profile (e.g. lack of convergence

due to test paralysis rates near 1.00) then STEPS 1-5 must be repeated

using the pooled data from the first and second experiment.

The test vaccine must satisfy the above criteria in order to pass. There will be no provision for

additional retesting. If, after retesting, the log odds ratio for the test vaccine exceeds L1 in any of

the above STEPS, it is deemed to have failed. If a pass/fail decision cannot be reached because of

problems with the reference (i.e. Paralysis rates <0.05 or >0.95, or no significant dose effect for

the reference) then the mouse test must be re-established before any further testing is carried out.

Acceptance and Rejection Limits – Starting Values and Updating Values In a laboratory establishing the current test, the absence of established historical reference data

means that limits cannot initially be set. In this instance the limits shown above are used, these

are in turn derived from the following historical values (obtained from collaborative studies) for

the WHO (SO+2) reference materials (1):


of 39

Females

Low Dose

Females

High Dose

Males

Low Dose

Males

High Dose

WHO (SO+2)/I

(Phase 5, 1.75 and 2.75)

0.162

(13/80)

0.785

(62/79)

0.133

(10/75)

0.705

(55/78)

WHO(SO+2)/II

(Phase 5, 5.0 and 6.0)

0.157

(30/191)

0.529

(101/191)

0.385

(74/192)

0.818

(0.157/192

WHO(SO+2)/III

(Phase 4, 3.5 and 4.5)

0.167

(34/204)

0.522

(106/203)

0.310

(64/205)

0.701

(141/201)

As each experiment is successfully completed within a laboratory (including experiments

completed during the training and implementation procedures) its reference vaccine paralysis

data is pooled with the above data and limits are updated following the procedure outlined below

Updating Acceptance and Rejection Limits

This section provides formulas which are used to calculate the acceptance limit (L1), the rejection

limit (L2) and the approximate probabilities of passing or failing vaccines. The symbols used in

the formulas are defined below:

i denotes the dose/gender group:

1= females low dose, 2= males low dose

3= females high dose, 4= males high dose

P1i = reference paralysis proportions in group i

P2i = test vaccine paralysis proportions in group i

Q1i = 1 - P1i

Q2i = 1 - P2i

Ri = Odds Ratio in group I

Li = Log Odds Ratio in group i

n i1 = sample size for reference in group i

n i2 = sample size for test in group i

The log odds ratio for subgroup ‘i’ is:

L Ri i= log( ) = logP Q

Q P

i i

i i

2 1

2 1

.

The common log odds ratio for the four groups is:

Lw L

w

i i

i

=∑

∑

with a standard error of

s ewi

. .=

∑

1


of 39

where

1 1 1

1 1 1 2 21 2w n P Q n P Qi i i i i i

= + .

L is approximately normally distributed. The probability that the estimated Log Odds Ratio in an

experiment will exceed a value of C is approximately:

Prob = Pr( Z > (C-L) / s.e.)

where Z is the standard normal deviate. The acceptance limit L1 was selected so that a test

vaccine which is equivalent to the reference will have a 0.95 probability of passing. That is,

L1 = 1.645 s.e.

where 1.645 is the 5% standard normal deviate . The rejection limit L2 was selected so that a test

vaccine which is equivalent to the reference will have a 0.01 probability of failing. That is,

L2 = 2.326 s.e..

For the purpose of calculating the starting limits, the historical reference paralysis rates were

substituted for the reference probabilities, P1i , in the above equations. For any specified odds

ratio, R, test vaccine paralysis proportions, P2i ,can be derived as follows.

Given the P i1 and a common odds ratio R we have

P

QR

P

Q

i

i

i

i

2

2

1

1

=

( )⇒ = −P RP

QPi

i

i

i2

1

1

21

⇒ =

+

P

RP

Q

RP

Q

i

i

i

i

i

2

1

1

1

1

1

These formulas were used to generate the acceptance/rejection limits shown in the table above,

and can be used to calculate probabilities of rejection or acceptance for any specified value of R.

The reference paralysis rates provided above were used as starting reference values for

laboratories which did not have an established reference database with the official methodology.


of 39

These reference values should be regularly updated. After each successful test with the reference

the results of the test are added to the historical database to obtain revised estimates of P1i . An

example of the calculation of the revised historical reference paralysis proportions after a first

experiment for Type 3 is:

Group Initial Proportions First Experiment Updated

Proportions

Females dose 3.5 0.167 (34/202) 2/16 0.164 (36/218)

Males dose 3.5 0.310 (63/203) 4/16 0.306 (67/219)

Females dose 4.5 0.522 (107/205) 9/16 0.525 (116/221)

Males dose 4.5 0.701 (141/201) 13/16 0.710 (154/217)

The revised proportions are then substituted into the above formulas to obtain revised limits for

this laboratory. For the above example the revised limits are L1=0.648 and L2= 0.917.

At the end of the Training and Implementation program (Appendices 2 and 3) a laboratory will

have garnered data for each serotype from at least 5 separate tests. These data should be used to

establish L1 and L2 values for the laboratory. These values should be updated with any

subsequent data generated. It is recommended that laboratories use the data from the last 10 tests

to recalculate values for L1 and L2.

Decision Model – Doses Used.

The values of L1 and L2 used for the pass/fail decision model outlined above are based on the

paralysis rates observed for the WHO (SO+2) reference. As the paralysis rate in the test is

dependent on the dose inoculated i.e. paralysis rates show a dose response across the range of

inoculum titres used in the tests there is a need to ensure that the doses used are within the

accepted limits of variability, as close as possible to the target doses and/or not significantly

different between test vaccine and WHO(SO+2).

Acceptance of Doses Used- Reference Calibration

It is recommended that the dilutions used in a test be assayed for titre at least twice, i.e. once

before use and once after use.

Data from each assay point should be pooled (with at least three replicates from each assay point)

and the pooled data should meet the following criteria:

• precision of ± 0.3 log10 CCID50/ml or better for the 95% confidence limits of the mean for

each dose used

• mean titre should be within ± 0.3 log10 CCID50 of the target dose.

Acceptance of Doses Used – Agreement with Doses of the WHO(SO+2) Reference

The doses used for the reference and for the test vaccine must be in agreement. Normal in-house

procedures for ensuring the validity of the assay must be followed.

The following criteria should be applied:

• precision of ± 0.3 log10 CCID50/ml or better for the 95% confidence limits of the mean for

each dose used


of 39

• mean titre of the in-house reference included in the assay should be within ± 0.3 log10

CCID50 of the established value.

• mean titre of the each dose of the WHO(SO+2) standard should be within ± 0.3 log10

CCID50 of the target dose

• mean titer of each vaccine dose should be within ± 0.3 log10 CCID50 of the equivalent

dose for the WHO(SO+2)

Worked example: In a type 3 test the titre of the WHO reference is found to be 6.7 and 5.7 log10

CCID50/ml for the 4.5 and 3.5 log10 CCID50/5µl doses respectively with 95% confidence limits of

0.2 and 0.3 log10 CCID50/ml. The titre of the in-house reference is found to be 8.2 log10

CCID50/ml compared to an assigned value of 8.4 log10 CCID50/ml. The mean titres for the vaccine

are 6.9 and 5.8 log10 CCID50/ml for the 4.5 and 3.5 log10 CCID50/5µl doses respectively. Using

the criteria above, these doses are acceptable:

• the 95% confidence intervals of the estimated titres are within the ± 0.3 log10 CCID50/ml

specified

• the mean titre of the in-house reference is within ± 0.3 log10 CCID50 of the assigned value

• the 4.5 and 3.5 log10 CCID50/5µl doses correspond to target titres of 6.8 and 5.8 log10

CCID50/ml and the titres of the WHO(SO+2)/III doses are within the specified limit (± 0.3

log10 CCID50) of these target titres

• the titres for the vaccines doses are both within the specified limit (± 0.3 log10 CCID50) of

the doses for the corresponding WHO(SO+2)/III doses.

Test dilutions should only be inoculated in mice if doses have been found acceptable. If the titre

of any of the vaccine doses used is found to be out of specification when titred after the test, a

repeat of the back-titration is possible and the results of both titrations could be pooled and

reassessed, but this should always be in agreement with the standard retest policy. If the titre of

any of the vaccine doses used is found to be outside of the limits noted above after retesting, a

statistical review of the data should be performed to decide whether the TgmNVT needs

repeating and new dilutions prepared.


of 39

Appendix 5: Monitoring of data from the Transgenic Mouse Neurovirulence Test

(TgmNVT)

The Transgenic Mouse Neurovirulence Test (TgmNVT) generates a large amount of data that

need to be monitored to ensure the performance of the test remains consistent and that pass/fail

decisions are consistent and reliable.

Each stage of the training and implementation process, as well as the subsequent testing program,

needs to be monitored and documented for both individual inoculators and organizations. It is

also possible that these data may be requested by the NCL or other relevant authorities.

The following provides an outline of the data that need to be monitored. Additional data may be

requested by the NCL.

Dye Inoculation

Records should be kept of the performance of each inoculator for each test carried out during the

training period (Appendix 2) and for any subsequent dye inoculations carried out e.g. those

carried out after a break in inoculations (see Appendix 2 – Maintenance of Competence)

Clinical scoring

A log of the clinical scoring should be kept for each observer. In the case of a long period of

inactivity (> 6 months) measures should be taken to reconfirm competence in observation skills

before a test scoring session is undertaken and these should be documented.

Titration of References

The TgmNVT relies on the accurate estimation of the titre of virus inocula used during the test

and on the match of these titres to the nominal doses used in the test. The titre of the titration

reference should be monitored to ensure the consistency of the assay.

Traumatic Paralysis

Records should be kept of the number animals that suffer from traumatic paralysis as a result of

the inoculation. If ≥5% of inoculated animals suffer traumatic paralysis, the inoculation technique

is inadequate and actions should be taken to improve performance.

Paralysis Rates for the WHO Reference

The paralysis rates for the WHO reference should be monitored for each dose and each inoculator

and records maintained.

LOR for vaccine batches

L-values should cover all tests conducted by the laboratory, but to detect variability between

operators, laboratories may wish to additionally calculate L-values for individual operators.

The correlation of LOR values and the results obtained for the MAPREC or other in vitro tests of

neurovirulence may additionally be monitored for each operator to increase consistency.


of 39

Authors

The first draft of the Standard Operating Procedure for the Neurovirulence Test of Types 1, 2, or

3 Live Poliomyelitis Vaccines (Oral) in Transgenic Mice Susceptible to Poliovirus (Version 2)

was written by Dr E. Dragunsky, U.S. Food and Drug Administration, Bethesda, Maryland, USA;

and Dr K. Karpinski, Ottawa, Ontario, Canada. Versions 3 to 5 were developed through a series

of workshops held with interested vaccine manufacturers, National Control Agencies, the

European Directorate for the Quality of Medicines (EDQM), and the World Health Organization

(WHO). Participants at these workshops included: Dr P. Christian, National Institute of

Biological Standards and Control (NIBSC), UK; Dr J. Martin, NIBSC, UK; Ms G. Cooper,

NIBSC, UK; Dr Ghazi Auda, NIBSC, UK; Dr E. Coppens, Sanofi Pasteur SA, Marcy L'Etoile,

France; Dr B. Descampe, GlaxoSmithKline Biologicals, Rixensart, Belgium; Dr L. Fiore, Public

Health Institute, Rome, Italy; Dr A. Heath, NIBSC, UK; Dr S. Lambert, World Health

Organization, Geneva, Switzerland; Dr M. Baca-Estrada, World Health Organization, Geneva,

Switzerland; Dr J. Fournier-Caruana, World Health Organization, Geneva, Switzerland; Dr A. Di

Lonardo, Public Health Institute, Rome, Italy; Dr A. Maes, Scientific Institute of Public Health,

Brussels, Belgium; Mr Graham Crossland, NIBSC, UK; Ms Lisa Pickard, NIBSC, UK; Dr C.

Milne, EDQM, Council of Europe, Strasbourg, France; Dr M. Pares, AFSSAPS, Lyon, France;

Dr T. Pasquali, Novartis, Siena, Italy; Dr F. Pelloquin, Sanofi Pasteur, Marcy L'Etoile, France; Dr

L. Perini, Novartis, Siena, Italy; Dr V. Pithon, AFSSAPS, Lyon, France; Dr I. Susanti, Biofarma,

Bandung, Indonesia; Dr G. Waeterloos, Scientific Institute of Public Health, Brussels, Belgium;

Mr. M. Janssen, Scientific Institute of Public Health, Brussels, Belgium; Dr R. Wagner, Paul

Ehrlich Institute, Langen, Germany, and Dr D. Wood, World Health Organization, Geneva,

Switzerland. Appreciation is expressed to Dr P. Minor, NIBSC, UK, and Dr C. Milne, EDQM,

Strasbourg, France for their expert review of the document.

The current version (Version 6) was prepared by Dr Catherine Milne, European Directorate for

the Quality of Medicines and Health Care, Council of Europe (EDQM), France and Dr Javier

Martin, National Institute for Biological Standards & Control (NIBSC) taking into consideration

requests and agreements at the Sixth NIBSC/EDQM Workshop on Training and Implementation

of the Transgenic Mouse Test for Oral Polio vaccine held at NIBSC,UK, between 10-12 October

2011, with representatives of vaccine manufacturers, National Control Laboratories, the

European Directorate for the Quality of Medicines (EDQM), and the World Health Organization

(WHO). Other participants at this workshop included (in alphabetical order): Dr Ghazi Auda,

(NIBSC), UK; Dr Elisa Brandini, Novartis, Italy; Dr Amélie Castiaux, GlaxoSmithKline

Biologicals (GSK), Belgium; Ms G. Cooper, NIBSC, UK; Dr E. Coppens, Sanofi Pasteur (SP),

France; Mr Graham Crossland, NIBSC, UK; Dr B. Descampe, GSK Belgium; Dr A. Di

Leonardo, Istitute Superiore Sanita (ISS), Italy; Ms Glynis Dunn, NIBSC, UK; Dr L. Fiore, ISS,

Italy; Dr J. Fournier-Caruana, WHO, Switzerland; Dr Andrea Gaggioli, ISS, Italy; Dr A. Heath,

NIBSC, UK; Mr. M. Janssen, SIPH, Belgium; Ms Jackie O’Brien, NIBSC, UK; Dr M. Pares,

Agence Francaise de Sécurité Sanitaire de Produits de Santé (AFSSAPS), France; Dr Tomasso

Pasquali, Novartis, Italy; Ms Lisa Pickard, NIBSC, UK; Dr V. Pithon, AFSSAPS, France; .Ms

Laura Stephens, NIBSC, UK; Stéphanie Van Brabant, GSK, Belgium; Dr Florence Vaudey, SP,

France; Dr Tiequn Zhou, WHO, Switzerland. Version 6 was subsequently reviewed during a

WHO consultation held on 27 -29 March 2012 attended by Dr Shinobu Abe, Japan Poliomyelitis

Research Institute (JPRI), Tokyo, Japan; Dr Maria Baca-Estrada, Biologics and Genetic

Therapies Directorate, Ottawa, Canada; Dr Wilfried A.M. Bakker, National Institute for Public

Health and the Environment (RIVM), Bilthoven, The Netherlands; Dr Jacqueline Fournier-

Caruana, Quality, Safety, and Standards, Immunization, Vaccines, and Biologicals,Family,


of 39

Women's and Children's Health, World Health Organization, Geneva, Switzerland; Mr

Bhupender Singh Chauhan, Bharat Biotech International Limited, Hyderabad, India; Dr

Konstantin Chumakov, Office of Vaccine Research and Review, Food & Drug Administration

Bethesda, USA; Dr Emmanuelle Coppens, Sanofi Pasteur, Marcy L’Etoile, France;Dr Michel

Duchêne,GSK Biologicals, Wavre, Belgium; Ms Glynis Dunn, National Institute for Biological

Standards and Control, South Mimms, UK; Dr Diana Felnerova, Crucell, Berne, Switzerland; Dr

Lucia Fiore, Istituto Superiore di Sanita, Roma, Italy; Mr José Bugarín González, Laboratorios de

Biológicos y Reactivos de México S.A. de C.V. México D.F., Mexico; Dr Martha Ayala

Gonzalez, Federal Commission for the Protection from Sanitary Risks (COFEPRIS), Ministry of

Health, Mexico; Prof Victor Grachev, Russian Academy of Medical Sciences (RAMS), Moscow

Region, Russian Federation; Ms Wang Hui, Tiantan Biological Products Co., Ltd , Beijing,

China; Mrs Teeranart Jivapaisarnpong, Department of Medical Sciences, Ministry of Public

Health, Bangkok, Thailand; Dr Ivana Knezevic, Quality, Safety, and Standards, Immunization,

Vaccines, and Biologicals, Family, Women's and Children's Health, World Health Organization,

Geneva, Switzerland; Dr Hiromasa Okayasu, Research, Policy and Product Development, World

Health Organization, Geneva, Switzerland; Dr Dede Kusmiaty, National Quality Control

Laboratory of Drug and Food, Ministry of Health, Jakarta, Indonesia; Dr Kazuhiko Katayama,

National Institute of Infectious Diseases (NIID), Tokyo, Japan; Dr Changgui Li, National

Institutes for Food and Drug Control (NIFDC), Beijing, P.R.China; Dr Javier Martin, National

Institute for Biological Standards and Control (NIBSC), South Mimms, UK; Dr Catherine Milne,

European Directorate for the Quality of Medicines and HealthCare (EDQM), Strasbourg, France;

Dr Rajiv Modi, Cadila Pharmaceuticals Limited, Ahmedabad, India; Ms Elisabeth Niogret,

Sanofi Pasteur, Marcy L'Etoile, France; Dr Le Van Phung, National Institute for Control of

Vaccine and Biologicals, Hanoi, Vietnam; Dr Virginie Pithon, Agence Française de Sécurité

Sanitaire des Produits de Santé (AFSSAPS), Lyon, France; Dr Alexandra Sinyugina, Federal

State Unitary Enterprise of Chumakov Institute of Poliomyelitis and Viral Encephalitides,

Russian Academy of Medical Sciences (RAMS), Moscow Region, Russian Federation; Dr

Roland Sutter, Research, Policy and Product Development, PEC/POL/RAP, World Health

Organization, Geneva, Switzerland; Mr Dori Ugiyadi, BioFarma, Bandung, Indonesia; Dr

Geneviève Waeterloos, Scientific Institute of Public Health, Brussels, Belgium; Dr Shudo

Yamazaki, Japan Poliomyelitis Research Institute, (JPRI), Tokyo, Japan; Mr Li Yi, Institute of

Medical Biology, Kunming, P.R. China; Dr TieQun Zhou, Quality, Safety and

Standards/Immunization, Vaccines and Biologicals/Family, Women's and Children's Health,

World Health Organization, Geneva, Switzerland.

Further changes were made, taking into consideration comments received during the public

consultation along with the WHO/BS/2012.2185, and following the review by the 63rd

Expert

Committee on Biological Standardization,15-19 October 2012, resulting in the present document.

References

1. Recommendations to Assure the Quality, Safety and Efficacy of Live Attenuated

Poliomyelitis Vaccine (oral). Proposed replacement of: TRS 904, Annex 1 and Addendum

TRS 910, Annex 1. Expert Committee on Biological Standardization 2012. World Health

Organization.

2. Dragunsky E, Nomura T, Karpinski K, Furesz J, Wood DJ, Pervikov Y, Abe S, Kurata T,

Vanloocke O, Karganova G, Taffs R, Health A, Ivshina A, and Levenbook I. Transgenic

mice as an alternative to monkeys for neurovirulence testing of live oral poliovirus


of 39

vaccine: validation by a WHO collaborative study. Bulletin of the World Health

Organization, 2003, 81:251-260.

3. Maintenance and distribution of transgenic mice susceptible to human viruses:

memorandum from a WHO meeting. Bulletin of the World Health Organization. 1993,

71:497-502.

4. Guidelines for the safe production and quality control of inactivated poliomyelitis vaccine

manufactured from wild polioviruses. In: WHO Expert Committee on Biological

Standardization. Fifty-third Report. (Addendum to the Recommendations for the

Production and Quality Control of Poliomyelitis Vaccine (Inactivated). Geneva. World

Health Organization, 2003. (WHO Technical Report Series, 910, 32-66,).

neurovirulence test of types 1, 2 or 3 live … · who sop for neurovirulence testing of oral polio...

Documents